DESCRIPTION (Applicant's Description): The long-term objective of the proposed research is the elucidation of the biological basis of particular aspects of human perception, namely, lightness and color constancy. Humans unconsciously process visual input in such a way as to extract context-independent information about objects (e.g., surface reflectance). This extraction occurs despite the fact that the context (e.g., the spectrum of the illumination) can drastically alter the properties of the light emanating from the objects. Understanding the various mechanisms employed for this extraction will facilitate development of artificial visual systems to function in their own right in robotic hardware as well as in aid to humans whose sight is impaired through accident, disease or developmental defect. The proposed research aims to test psychophysical linking hypotheses tying perception to events in the eye. Specifically, electroretinographic (ERG) recordings will be made during the performance of a psychophysical task. By combining data from human subjective reports with data about physiology in the retina, it will be possible to assess the degree to which particular forms of perceptual constancy are attributable to constancies in the signals available at the retina. To achieve these goals, subjects will compare flickering patches displayed on a computer monitor as the computer acquires ERG data from the subjects' eyes. The experimental design will determine conditions under which two flickering stimuli appear the same even though they are physically distinct. Instances in which components of the ERG signal are the same when subjects report the perceptions are the same will indicate that the perceptual constancy is related to constancy in the signal from the retina. Conversely, perceptual constancy in the face of differences in ERG signals will imply that the constancy is a result of higher-level processing of retinal signals.